[Linux-ha-dev] Announcing crmsh release 2.1.7

[Kristoffer Grönlund]

Hello everyone!

Today I are proud to announce the release of `crmsh` version 2.1.7!
The major new thing in this release is a backports of the events-based
alerts support from the 2.3 branch.

Big thanks to Hideo Yamauchi for his patience and testing of the
alerts backport.

This time, the list of changes is small enough that I can add it right
here:

- high: parse: Backport of event-driven alerts parser (#150)
- high: hb_report: Don't collect logs from journalctl if -M is set (bsc#990025)
- high: hb_report: Skip lines without timestamps in log correctly (bsc#989810)
- high: constants: Add maintenance to set of known attributes (bsc#981659)
- high: utils: Avoid deadlock if DC changes during idle wait (bsc#978480)
- medium: scripts: no-quorum-policy=ignore is deprecated (bsc#981056)
- low: cibconfig: Don't mix up CLI name with XML tag

You can also get the list of changes from the changelog:

* https://github.com/ClusterLabs/crmsh/blob/2.1.7/ChangeLog

Right now, I don't have a set of pre-built rpm packages for Linux
distributions ready, but I am going to make this available soon. This
is in particular for centOS 6.x which still relies on Python 2.6
support which makes running the later releases there more
difficult. These packages will most likely appear as a subrepository
here (more details coming soon):

* http://download.opensuse.org/repositories/network:/ha-clustering:/Stable/

Archives of the tagged release:

* https://github.com/ClusterLabs/crmsh/archive/2.1.7.tar.gz
* https://github.com/ClusterLabs/crmsh/archive/2.1.7.zip


Thank you,

Kristoffer

[Darren Thompson]

Team

good work on the new version, appreciated.

Just a quick question:

If "scripts: no-quorum-policy=ignore" is becoming depreciated, how are we
to manage two node (e.g. test) clusters that require this work around since
quorum state on a single node is an odd state.

Regards





Darren Thompson

Professional Services Engineer / Consultant



Level 3, 60 City Road

Southgate, VIC 3006

Mb: 0400 640 414

Mail: ***@akurit.com.au <***@akurit.com.au>
Web: www.akurit.com.au

Post by Kristoffer Grönlund
Hello everyone!
Today I are proud to announce the release of `crmsh` version 2.1.7!
The major new thing in this release is a backports of the events-based
alerts support from the 2.3 branch.
Big thanks to Hideo Yamauchi for his patience and testing of the
alerts backport.
This time, the list of changes is small enough that I can add it right
- high: parse: Backport of event-driven alerts parser (#150)
- high: hb_report: Don't collect logs from journalctl if -M is set (bsc#990025)
- high: hb_report: Skip lines without timestamps in log correctly (bsc#989810)
- high: constants: Add maintenance to set of known attributes (bsc#981659)
- high: utils: Avoid deadlock if DC changes during idle wait (bsc#978480)
- medium: scripts: no-quorum-policy=ignore is deprecated (bsc#981056)
- low: cibconfig: Don't mix up CLI name with XML tag
* https://github.com/ClusterLabs/crmsh/blob/2.1.7/ChangeLog
Right now, I don't have a set of pre-built rpm packages for Linux
distributions ready, but I am going to make this available soon. This
is in particular for centOS 6.x which still relies on Python 2.6
support which makes running the later releases there more
difficult. These packages will most likely appear as a subrepository
* http://download.opensuse.org/repositories/network:/ha-
clustering:/Stable/
* https://github.com/ClusterLabs/crmsh/archive/2.1.7.tar.gz
* https://github.com/ClusterLabs/crmsh/archive/2.1.7.zip
Thank you,
Kristoffer
--
// Kristoffer Grönlund
_______________________________________________________
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/

[Kristoffer Grönlund]

Post by Darren Thompson
If "scripts: no-quorum-policy=ignore" is becoming depreciated, how are we
to manage two node (e.g. test) clusters that require this work around since
quorum state on a single node is an odd state.

Hi Darren,

There are better mechanisms in corosync and Pacemaker for handling two
node clusters now while still maintaining quorum.

In corosync 2, we have the two_node: 1 setting for votequorum, which
ensures that a two node cluster doesn't suffer split brain (fencing is
required for this to work properly).

There is an explanation for how this works here:

http://people.redhat.com/ccaulfie/docs/Votequorum_Intro.pdf

Somewhat related, there used to be the start-delay meta parameter which
could be set for example for sbd stonith resources, to make a
double-fencing scenario less likely. This has now been replaced by the
pcmk_delay_max parameter. For an example of how to use this, see this
pull request for sbd:

https://github.com/ClusterLabs/sbd/pull/15/commits/ca2fba836eab169f0c8cacf7f3757c0485bcfef8

Cheers,
Kristoffer

[Darren Thompson]

Thank you.

Comprehensively answered.

Post by Kristoffer Grönlund

Post by Darren Thompson
If "scripts: no-quorum-policy=ignore" is becoming depreciated, how are we
to manage two node (e.g. test) clusters that require this work around

since

Post by Darren Thompson
quorum state on a single node is an odd state.

Hi Darren,
There are better mechanisms in corosync and Pacemaker for handling two
node clusters now while still maintaining quorum.
In corosync 2, we have the two_node: 1 setting for votequorum, which
ensures that a two node cluster doesn't suffer split brain (fencing is
required for this to work properly).
http://people.redhat.com/ccaulfie/docs/Votequorum_Intro.pdf
Somewhat related, there used to be the start-delay meta parameter which
could be set for example for sbd stonith resources, to make a
double-fencing scenario less likely. This has now been replaced by the
pcmk_delay_max parameter. For an example of how to use this, see this
https://github.com/ClusterLabs/sbd/pull/15/commits/
ca2fba836eab169f0c8cacf7f3757c0485bcfef8
Cheers,
Kristoffer
--
// Kristoffer Grönlund

[Kristoffer Grönlund]

Hi,

Post by Darren Thompson
If "scripts: no-quorum-policy=ignore" is becoming depreciated

Are there any plans to get rid of this option?
Am I missing something?

The above is talking about crmsh cluster configuration scripts, not core
Pacemaker. As far as I know, no-quorum-policy=ignore is not being
deprecated in Pacemaker.

However, it is no longer the recommended configuration for two node
clusters.

PS: this option is very useful (vital) to me. And "two_node" option won't
replace it.

Out of curiosity: What do you use it for, where the two_node option is
not sufficient?

Cheers,
Kristoffer

[Kristoffer Grönlund]

Post by Kristoffer Grönlund

Out of curiosity: What do you use it for, where the two_node option is not

sufficient?
Alongside with starting the cluster with two nodes I need that possibility
of starting the cluster with only one node.
"two_node" option doesn't provide that.
Thank you,
Kostia

That's true, for that you'll need to set no-quorum-policy=ignore. But,
to clarify again, no-quorum-policy=ignore is still there, it is just not
the recommended default configuration for a two node cluster.